88 ALASKAN GLACIER STUDIES in detail, and although in a region of deposit, where the normal change should be an outward growth, it had receded several yards and a gravel knoll that stood there in 1905 had been nearly consumed. The site of our camp of 1905, which was on the beach at the base of this knoll had disappeared also. Slumping of the alluvial fan back of the beach was also indicated by the fact that, although this area had for four years received the drainage of the large stream that flowed from the western margin of Galiano Glacier, its bottom had not yet been raised above the reach of the tide. Indeed, it seemed even " r> lower than it was in 1905. Out on the alluvial fan-moraine area in front of the visible ice bulb of the Galiano there had been much change in detail, but principally of two kinds. In the first place there had been a great increase in the amount and density of vegetation. On the higher and more extensive morainic knolls there were now alder thickets through which in places one found it difficult to travel, whereas in 1905 there were no such areas. On the lower knolls there was less noticeable increase in vegetation, but we had good evidence of increase in one place when we attempted to find the site of one of our 1905 photographs and, in looking for the foreground stones, found them partly hidden by alders, whereas in 1905 there were only annual and perennial plants and a few scattered alders one or two years old. The entire foreground (compare vegetation near stones marked A, B, C, and D in PL XXXIV) was completely altered by growth of vegetation. A second noteworthy change is that of deposit among the morainic hillocks. The process of filling the depressions, and thus of submerging the hillocks beneath gravel deposit, had proceeded so far as to be noticeable in some places; but we had no data for determining whether at the same time there had been a lowering of the hillocks by the melting out of the underlying ice, though this must inevitably have been the case, to some extent at least, for large quantities of cold water still issued from the moraine and this had without doubt been continuing during the intervening summers. The water that escaped represented ice that was lost by melting, and this loss must have caused a lowering of the moraine. It is natural that the anastomosing glacial streams which are building an outwash gravel plain between the moraines should change their courses as they build one portion higher than another. A number of such changes had occurred, but one was specially noteworthy. In 1905 and 1906 the Galiano stream flowed over an alluvial fan which it had built between the visible glacier front and the first range of low morainic bills a quarter of a mile to the south of it. Over this fan the stream wandered in branching and ever-shifting courses, and at times a branch swung over against the moraine which it had trimmed back into a fairly steeply-sloping bank. On the morning of July 28, 1909, the Galiano stream was flowing as usual, with three large distributaries and several smaller ones, all of which were easily forded, though the northern one was fairly deep and swift; but a small branch had found its way southward through a gap in the moraine. When we returned at night our progress was barred by a single torrent which outflowed through the gap, and which was crossed with difficulty. All the other stream courses were abandoned, and for the time being the entire lower portion of the alluvial fan received no water or sediment. The new course proved so favorable that it was maintained as long as we remained there (until August 8), and that it was likely to be followed for some time was indicated by the fact that the torrent had entrenched itself in a newly-formed gravel gorge which was extending up-stream.